Project
DK‑Future estimates how land use in Denmark will change under compound climate impacts such as flooding, drought and coastal change.
We develop probabilistic geospatial machine learning (GeoML) models that account for spatial structure in land use. The models are conditioned on earth observation, climate projections, socio‑economic indicators and land‑use legislation, and output a predictive distribution over land‑use scenarios for each forecast horizon.
- Funding
- Villum Synergy
- Grant
- DKK 4.48 million
- Period
- 2026–2028
- Host
- Aalborg University, Dept. of Computer Science and Dept. of Sustainability and Planning
Research
-
Probabilistic GeoML
Land‑use models that account for proximity effects and spatial autocorrelation.
-
Uncertainty quantification
Calibrated predictive distributions, including the probability of low‑likelihood, high‑impact scenarios.
-
Compound climate impacts
Land‑use response to flooding, drought and coastal change under climate projections.
-
Open science
Code, datasets and trained models will be released openly.
Publications
No publications yet.